I'd like to know what conceivable inputs the crew might have received that would make them decide to have everyone sit tight in a burning and generously fuelled airplane.
You can bet that being blind sided by this fire after a successful but tiring night mission abort and successful landing counted for at least 30 seconds of delay in recognition of the fire. Then follows confusion about which procedures to implement. By then, the fire department is there. I expect the CVR will pretty much confirm this pathway to a passive decision.
In an earlier post, I mentioned the significance of the fuel odor reported in the cabin. Since the pressurization system on the triple 7 takes its air from the engines, that implies fuel may be getting into the bleed air. How? Leakage from an engine lube system that has become contaminated with fuel. The point of this comment is to suggest that the EICAS system should not be the exclusive input to your decision making process regarding status of aircraft systems.
I'm in substantial agreement with the following analysis by Whinging Tinny :
Main Fuel/Oil Heat Exchanger failure (Known fault on GE engines)
Higher presssure fuel enters the oil system and circulates around the oil system - hence the oil system parameters changing.
Fuel/oil vapour/mist vents out the central vent system and to atmosphere.
Due to cold stream airflow/forward speed of the aircraft nothing happens.
Flaps down on approach and lowering airspeed, venting fuel/oil vapour impinges on the lower wing surface.
Selecting reverse thrust plus decaying forward speed due to landing, no more airflow to draw the vapour away.
Engine central vent system vapour ignites and spreads to the contaminated lower surface.
Even after engine shutdown, system still pressurised and venting adding to problem.
Where we differ is that I expect the accident investigation will find a cascaded failure of the vent system/oil system in/adjacent to the engine pylon. The pattern of fire indicates to me that the pylon became awash in fuel and transmitted fuel into the leading and trailing edge flap spaces under the influence of local airflow which I expect will run from wing root to tip. The amount of fuel would likely be in excess of 5000 lb over the course the the 2+ hour return trip, with the bulk of the fuel being lost overboard. On landing, any fuel going down the core will cause an immediate compressor stall and start things burning. The unusual pattern of this fire and the external physical design and location of the engine pylon lead me in this direction. Note that in the first picture below, there is no leading edge fire inboard of the engine pylon.
